From: "James Moritz" <J.R.Moritz@herts.ac.uk>
To: rsgb_lf_group@blacksheep.org
Date: Tue, 24 Feb 2004 12:26:46 -0000
Organization: University of Hertfordshire
Message-ID: <000001c3fad1$785a5280$fce8c593@RD40004>
MIME-Version: 1.0
Importance: Normal
In-reply-to: <HNEGLMNKFOEMMNHIFAPIAEOACDAA.g4wgt@talkgas.net>
Subject: LF: RE: RF Current meter
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 8bit
Sender: <owner-rsgb_lf_group@blacksheep.org>
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Dear Gary, LF Group,

The RF ammeter circuit in the LF handbook and on K0LR's web page uses a
current transformer to develop a voltage of a few volts across a load
resistor when the appropriate RF current is flowing in the primary. This
voltage is then measured by a simple diode voltmeter to indicate the
current. Sounds pretty simple then, but some pitfalls are possible with
an RF ammeter of this type...

The first thing is that, in order to obtain a reading that is not highly
dependent on the ferrite properties or the operating frequency, the
inductive reactance of the secondary winding must be much larger than
the load resistance. E.g., in the case of the design in the handbook,
the load resistor is 470ohms, and at 136k a secondary inductance of
several mH would be desirable. To obtain this with the 50turn winding of
the handbook design, the specific inductance of the ferrite core used
must be a few thousand nH/turn. This is readily achieved by high
permeability ferrite cores (relative permeability in the thousands)- HF
ferrite materials, or iron dust materials have too low permeability.
However, if such a core is split to make a clamp-on meter, it is
essential that the two halves meet together very accurately - even a
tiny, practically invisible, air gap will result in a drastic reduction
in inductance. The winding inductance has a shunting effect, so if the
inductance is too low, the meter will read low, decreasing in proportion
to frequency below some cut-off frequency (1/(2pi*L/R))- which may be
what is happening in Gary's case.

The second thing is the response of the voltmeter. The diode voltmeter
will be non-linear, reading low at low voltages due to the forward drop
of the diode. Increasing the RF load resistor will increase the voltage
for a given current, and so improve voltmeter accuracy, but will require
higher current transformer secondary inductance . The minimum errors at
low frequency will occur with lower values of load resistor. The
accuracy of the diode voltmeter will be improved by increasing the DC
resistance of the load, so it would be best to use the lowest-current
meter movement available. This also means that for Gary to extend the
range of his meter upwards, it would be better to reduce the load
resistor connected across the current transformer secondary, rather than
shunt the meter movement. Shunting the meter movement will increase the
output current required from the rectifier, increasing the error due to
the voltage drop across the diode, again tending to make the meter read
low.

...So, as usual, simple circuits have more to them than at first it
seems. A slightly different way of doing the same job is to dispense
with the RF load resistor, and feed all the RF current from the current
transformer through a bridge rectifier, and measure it with a
low-impedance DC ammeter. This makes the voltage across the current
transformer secondary as low as possible (the rectifier voltage drop,
plus the DC drop across the ammeter), minimising the shunting effect of
the winding inductance. Also, the accuracy and linearity is pretty good
even at low currents with this circuit, because the current transformer
secondary voltage is always just what it needs to be to drive the
rectifier. I used this approach for my LF tuning meter design to produce
an ammeter with switchable 2A/6A FSD - see
http://www.picks.plus.com/software/LFtunemeter.pdf on G3YXM's web pages.
The ammeter part is T3, D7-D10 and the associated resistors and meter,
and could be used by itself, of course. 

Cheers, Jim Moritz
73 de M0BMU